VMware has upgraded its application management software Hyperic claiming to offer a way for administrators to fix application problems more quickly.
Hyperic 4.4, developed by VMware company SpringSource, maintains a continually updated inventory of vSphere ESXi and ESX hosts, allowing administrators to identify and correct application performance problems as they occur.
According to SpringSource's Hyperic product markting manager, Al Sargent, the software has been aimed specifically at the applications operations team within an IT organisation. "Typically this group is distinct from the team responsible for implementing virtual infrastructure, which is why they haven't had good insight into the performance of vSphere hosts."
The software offers several changes for previous versions. "For the first time," said Sargent. "Hyperic 4.4 server now connects to VMware vCenter, essentially the nerve center of a data centre's virtual infrastructure, in order to pull information about vSphere hosts (ESXi and ESX instances), each of which has multiple VMs running on it. By doing this, Hyperic gains a 360 degree view of all virtualised application infrastructure, and provides sysadmins with rapid diagnosis of virtualised application performance problems and maintenance of the application infrastructure inventory."
Sargent added that there had been a marked improvement in the way that Hyperic could be used to isolate performance problems. He said that before the new software, application administrators were able to view performance metrics for a particular virtual machine, its guest operating system, and any apps running on that guest OS. "But Sargent said, the administrators could organise those VMs by vSphere ESXi/ESX host, or view the performance metrics for those vSphere hosts.
"Because of this, application admins had a visibility gap that prevented them from seeing a range of problems, such as overloaded vSphere hosts that had too many VMs running on them," Sargent said. "If you only look at the VM layer, but not the ESXi/ESX layer, you cannot see this pattern over host overutilization, and you can't diagnose your app performance problem."
He added that Hyperic could solve this problem by enabling administrators to group by host. "So, back to our overloaded vSphere host example, an app admin can now see that all VMs residing on a particular vSphere host are running slowly, and use that information to conclude that they need to move VMs off that host and onto one that's not as overtaxed. You have the visibility you need to solve a performance problem faster."